XXI. Mineralogical Notes. By EDWARD J. CHAPMAN, Professor of Mineralogy in University College, London*.

UNDER

NDER the above title, it is the intention of the writer to offer from time to time a series of remarks and investigations on subjects relating to mineralogy.

(1.) Crednerite.—If we allow the isomorphism, atom for atom, of silica and alumina, and through alumina, of silica and the sesquioxides generally+, Rammelsberg's crednerite-the mangankupfererz of Credner-may be admitted into the augite group. The cleavage form of crednerite is certainly monoclinic; and the angles, so far at least as they can be estimated in the specimens hitherto obtained, assimilate to those of the augite prism. The minerals of the augite type have the general formula 3RO, 2Si O3, which, with the substitution of sesquioxide of manganese for silica, is exactly that of crednerite as deduced by Rammelsberg, viz. 3(CuO, BaO), 2Mn2 03. A proof of the isomorphism of Si O3 and Mn2O3 is afforded by the manganese garnets and spinels, and more particularly as belonging to a system of variable angles, by Vesuvian and Hausmannite.

(2.) Helvine.-The helvine in the classification of Mohs bears the name of tetrahedral garnet; and with the spinel and garnet group it must in fact be placed, unless it stand alone, for to no other type amongst the silicates and their isomorphs can it be referred. On the supposition that Si Q3 and Mn2 03 are isomorphous, and that sulphur and oxygen are equally so, the atomic constitution of the helvine falls into the common garnet formula, r+R. The composition of helvine, for instance, as usually represented, (Be2O3, Fe2O3), SiO3 + 2MnO, SiO3+ MnS. This may be reduced into three atoms of (MnO, MnS), and three atoms of (R2 O3, Si O3); or into equal atoms of base and acid, r+R.

=

(3.) False cleavage in Garnet.-A small garnet in the author's possession, from the Zillerthal, exhibits a peculiar and interesting example of false cleavage in relation to the crystalline structure of metamorphic rocks. This garnet is a combination of a trapezohedron, 10, with the rhombic dodecahedron; and entirely through its mass run lines of false cleavage, parallel to the cleavage planes of the mica-slate in which it is imbedded.

* Communicated by the Author.

† See a paper by the Author on the isomorphous relations of silica and alumina, in the report of the British Association for 1850.

(4.) Phenacite and Beryl.-Phenacite was at first mistaken for quartz, hence the derivation of its name; it is, however, very closely related to that substance. Both quartz and phenacite are hemihexagonal: in the former, the relative length of the vertical axis = 1·095; in the latter, 0.5471; so that the common phenacite rhombohedron R compared to the quartz form as unity. In beryl, again, from the two triaxial pyramids occasionally present in that mineral, we obtain for the relative length of the vertical axis in the protaxial form, the values 0·9968 and 0.4980. The first is in close accordance with that of the protaxial form of quartz.

=

Phenacite and beryl may therefore be considered members of the quartz group, in which, if glucina be looked upon as a sesquioxide, there can be no difficulty in placing them. On the other hand, if the formula of glucina be written BeO, the isomorphism of 3BeO with Si O3, and the sesquioxide isomorphs of the latter, must be allowed. Of the other glucina compounds, the chrysoberyl may represent a trimetric, and the euclase a monoclinic quartz. With the former are associated staurolite, andalusite, topaz, &c. Helvine, as shown in note 2, belongs to the garnet type.

(5.) Sphene and Epidote.-If sphene constitute not a type of its own, the only group to which it can be referred is that of the epidote series. In epidote, the general formula—uniting the Si 03 and R2 03-becomes 73 R5. In sphene, the number of atoms =3CaO, 3TiO2, 2Si O3; and by uniting the acids, we obtain equally with epidote the formula 3 R5.

That silica and the sesquioxides are at times isomorphous with titanic and stannic acid, we have evidence in the isomorphism, on the one hand, of idocrase, hausmannite and anatase; and, on the other hand, of zircon, rutile and cassiterite. All of these forms belong to the dimetric system, but their axial relations separate them into two distinct groups.

(6.) Chlorite spar and Chloritoid.-These minerals, which closely resemble each other, may be looked upon as allied to epidote. Chlorite spar is, in fact, an iron epidote. Erdmann's analysis gives in atoms 3FeO, 3A12 03, 2Si Ô3, = r3R5.

The chloritoid, according to Bonsdorff's analysis, contains 3RO, 2A12 O3, 2Si O3, 3H2O. If we admit that three atoms of water may replace one atom of silica, or of alumina, these numbers produce, as before, the formula R5. It may be remarked in support of the above suggestion, that Rammelsberg considers one of the iolite metamorphs-the praseolite-to be an iolite in which one atom of silica has given place to three atoms of water.

(7.) Wichtyne. Minerals of the Epidote Type.-Laurent's wichtyne, the wichtisite of Hausmann, appears also to belong to the epidote type-if it be not actually an altered variety of epidote. It yields, however, by Laurent's analysis, 3RÓ, R2 Ó3, 4Si O3, whilst the epidote contains 3RO, 2R2 03, 3Si O3.

The epidote type may thus consist of the following minerals:allanite or orthite (including bagrationite, &c.), gadolinite, epidote, sphene, wichtyne, chloritoid and chlorite spar. The allanite and epidote are strictly isomorphous, but their chemical formulæ are by no means alike. The former contains, in atoms, 3RO, Al2 03, 2SiO3; the latter, 3RO, 2R2 O3, 3Si O3. Isomorphism, therefore, is no proof of kindred composition, or some extended hypothesis must be adopted to meet the above case. If we assume that 3r=1R-id est, that one atom of silica, or of a sesquioxide, three atoms of RO-the difficulty vanishes, and the two formulæ enter of course under one common term. This hypothesis is necessarily at present a purely gratuitous one, admitting, in fact, of the widest licence, and consequently of the widest abuse; but unless some hypothesis of the kind be, at least provisionally, adopted, we cannot retain our existing formulæ and effect at the same time a satisfactory distribution of minerals. Every fresh observation shows the insufficiency, for instanceeven if convenience plead for its retention-of the division of the silicates according to the oxygen relations of their so-called bases. Few mineralogists will now disallow the propriety of placing truly vicarious or isomorphous compounds under the same common type; but the difficulty lies in the legitimate employment of heteromerous isomorphism as a classification-element. That heteromeroisomorphous compounds should in some cases be placed together and in others be kept distinct, is, however, sufficiently evident; the grounds of union or separation constitute, therefore, the question at issue. Besides crystallization characters, three other elements should here be looked to;-first, the general chemical nature of the substance; secondly, its other physical characters; and thirdly, its circumstances of occurrence. On these data I would place phenacite and beryl with quartz, but not quartz with chabasite; acmite and augite, again, together, but not augite with borax. Numerous other examples will readily occur to those conversant with the subject.

(8.) Chrome Tourmalines.-Many of the Siberian tourmalines contain a small amount of chromium, probably as Cr2 03. These specimens are generally in acicular groups, and of an extremely fine green colour. At first sight they might be mistaken for actynolite; and, indeed, a specimen which I examined had a label attached to it bearing that name. H=70; sp. gr. =3.181. Fusible.

(9.) Detection of Manganese in Limestone Rocks.-A considerable number of limestone rocks contain a minute proportion of carbonate of manganese. I have noticed this more particularly in the darker magnesian limestones of the Permian epoch, but also in various other limestones of different ages and from different localities. The common blowpipe-test--fusion with carbonate of soda-generally of so delicate an appreciation, here fails, even on the addition of nitre, to point out the presence of manganese. This is owing to the insolubility of the limestone in the carbonate of soda. If, however, a very small quantity of borax be added, so as to attack and dissolve a portion of the mass, the well-known greenish-blue enamel is quickly produced.

(10.) Barytine.-The crystals of barytine (BaO, SO3) from the fuller's-earth pits in the greensand formation of Nutfield, near Bletchingly in Surrey, possess the general configuration of the Cumberland, Schemnitz, and other crystals in which the basal form P (OP of Naumann) predominates. The forms usually present in the Nutfield crystals are the prismatic forms P and D, and the diaxial pyramidal forms A, A, and E. L, A, and A are occasionally seen, as also the triaxial forms O, 40; but the latter are in general very minute.

In Naumann's notation, P = OP, L=∞ P∞, D=∞ P, A=P∞, E=Р∞, O=P.

D:D=101° 42'; P: A=121° 46'; P:A=141° 4' 30"; P: A 158° 0' 30"; P: A 162° 6'; P: E 127° 15'; P: O=115° 39'. Axes: V=1·315; T=1; F=0·8141. Viewed in regard to their general configuration, the crystals. of sulphate of baryta fall into six groups. In the following tabular arrangement, the vertical axis is denoted by V, the frontal by F, and the transverse (right and left axis) by T.

Group 1, with P predominating.-The crystals of this group have a more or less flattened or tabular appearance. Localities: Felsobanya, Schemnitz, Freiberg, Cumberland, Nutfield, &c.

Group 2, with A predominating.-Crystals elongated parallel to axis T. Puy-de-dome, Marienberg, Przibram, &c. A combination frequently met with in trap and volcanic districts.

Group 3, with E predominating.-Crystals elongated parallel to axis F. Puy-de-dome, &c. This configuration is very rare. Group 4, with P and E predominating conjointly.—These crystals are also elongated in general along the frontal axis, and are much more common than the above. Mies in Bohemia, Freiberg, Baden, Lancashire, Nutfield, Cheshire in Connecticut, &c. Group 5, with A and E predominating. These forms produce an irregular octahedron, generally elongated along the axis F. Puy-de-dome, &c.

Group 6, with D predominating.-Crystals elongated vertically. A very rare configuration principally exhibited by a few crystals from Siberia, and from the Siebenbirgen district.

The most common forms of sulphate of baryta are P, A, and E. D, O, and L are also of frequent occurrence, but they are comparatively of small size. L is more common than the back and front monaxial form M: the two are not often found in the same combination.

XXII. On the Electrical Properties of Flame. By H. BUFF, Professor of Natural Philosophy in the University of Giessen*.

PROFE

ROFESSOR BUFF commences his memoir with a review of the divergent notions at present existing as to the electricity of flame; Becquerel finds electric opposition in all directions in flame, which depends upon the difference of the temperature of the metals immersed in it; Pouillet recognises a motion of electricity only from the interior to the exterior, and hence also from the base to the summit of the flame; according to Hankelt, however, the motion of the electric fluid, at least in flames obtained by the ignition of spirit, is exactly opposite, and independent of the temperature of the immersed conductor. To solve these contradictions was the object of the present investigation.

Two small strips of platinum were introduced into a glass tube closed at one end; they were separated by an interval of 1.5 line of air. The air within the tube could not be heated to a degree sufficient to permit the electricity of two of Daniell's cells to pass through it. When the glass became soft by heating, and both pieces of platinum were permitted to touch it, a strong deflection of the needle of a galvanometer was the consequence.

A porcelain tube two feet long and six lines wide was encompassed with glowing coals, and air was drawn slowly through it; this air could not be heated so as to allow the passage of the electricity from the source above mentioned, although the two platinum wires sunk in the air were less than a line apart, and were glowing red.

A metal web was placed over the flame of a spirit-lamp; the flame did not pass through; over the web the platinum strips were held a line apart-there was no passage of electricity.

The galvanometer used in these experiments was extremely sensitive. When two persons who were connected simply by the

* Annalen der Chemie und Pharmacie, vol. lxxx. S. 1.

† Phil. Mag., S. 4. vol. ii. p. 542.

Phil. Mag. S. 4. Vol. 3. No. 16. Feb. 1852.

L